Unless an additional extension is granted, the Boeing 737 MAX 7 and MAX 10 cannot be certified after 2022-12-31 without inclusion of an engine-indicating and crew-alerting system (EICAS), which would require redesign of the flight deck and possibly invalidate the common type certificate which allows pilots to operate different 737 models without retraining. It was expected that both planes would be certified before this deadline, but delays due to the grounding of the 737 MAX after crashes and COVID-19 have caused certification to slip, according to Boeing, to early 2023 for the MAX 7 and early 2024 for the MAX 10. Unless U.S. regulators grant another extension, Boeing is once again backed into a corner with these models.
Sounds like time for Boeing execs. to invest in Hunter Biden paintings.
Not to worry. Boeing is undoubtedly making lots of money shipping weapons (gifts from the US taxpayer) to the Kiev Kleptocrats – weapons which will someday show up far from the Ukraine. And the Western sanctions on energy from Russia are working their way through to higher airline fuel costs and thus reduced travel and reduced demand for new airplanes, whether certified or not. All is proceeding according to plan!
While on the software process board for SAIC circa 1988, I was managing the development of control software for an automated inspection system. We were handling ordnance from fuses (x-ray) to 10" shells (gamma). Delivery of this system to the ordnance factory complex in Burlington, IA was amusing. The gate through the earth berm surrounding the factory greeted us with stats how how long it had been since anyone was killed, injured, etc.
Now, I’ll admit, mere ordnance inspection isn’t as critical as passenger air transport software in terms of “fatal” bugs, but it did give me a feel for the general area of concern. And I guess it was a combination of working on that system while on SAIC’s software process board, and taking a course in neural networks up the hill at UCSD that gave me an intuitive “feel” for the value of Occam’s Razor in algorithms. It was around this time that I made the mistake of not taking Ray Ozzie up on his offer to join his crew at Iris. I thought neural networks were going to take off and was thinking seriously about how generalization accuracy in neural networks arises from parsimony.
Later, when Ray took over from Gates at MS, I suggested a precursor idea to the Hutter Prize:
Hold an internal MS contest, funded to on the order of a billion dollars, for incremental reductions in the size of a binary that would expand into an installation that ran and passed all of the MS tests. This idea did not, get very far.
But it still strikes me as the right strategy for any large entity that is serious about reducing the complexity of their software suite to the point that the philosophy of the system is disciplined, bugs are factored out and the cyber-attack surface is reduced.
The main problem, as I see it, is not the RoI since substantial payout is almost certain to buy substantial increase in software quality.
The main problem is that the cyberwar has already been lost: To India.